Political conflict in Bismarck’s Germany: An analysis of parliamentary voting, 1867-1890 Frank M. Häge 1 Department of Politics and Public Administration University of Limerick Abstract Imperial Germany is a prominent historical case in the study of Western Europe’s political development. This paper investigates the number and content of political conflict dimensions from the foundation of the modern German state in 1866 to the end of Bismarck’s reign as Chancellor in 1890. Methodologically, it applies dimension-reducing statistical methods to a novel dataset of content-coded parliamentary roll call votes. The analysis suggests that the emergence of the Catholic Centre Party in 1871 permanently transformed the conflict space from a single liberal-conservative divide to a two-dimensional space that distinguished positions on socio-economic issues and regime matters, respectively. The fact that positions on redistributive and regime issues were not aligned implies that theories stressing economic inequality as a driver for regime change are of limited applicability. Instead, the case of Imperial Germany highlights the importance of cross-cutting non-economic societal cleavages and the role of societal and political organisations in drawing attention to and perpetuating these divisions. Keywords Dimensionality, political space, roll call votes, Reichstag, Imperial Germany 1 Email: [email protected]; web: www.frankhaege.eu. brought to you by CORE View metadata, citation and similar papers at core.ac.uk provided by University of Limerick Institutional Repository
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
Political conflict in Bismarck’s Germany: An analysis of
parliamentary voting, 1867-1890
Frank M. Häge1
Department of Politics and Public Administration
University of Limerick
Abstract
Imperial Germany is a prominent historical case in the study of Western Europe’s political
development. This paper investigates the number and content of political conflict dimensions
from the foundation of the modern German state in 1866 to the end of Bismarck’s reign as
Chancellor in 1890. Methodologically, it applies dimension-reducing statistical methods to a
novel dataset of content-coded parliamentary roll call votes. The analysis suggests that the
emergence of the Catholic Centre Party in 1871 permanently transformed the conflict space
from a single liberal-conservative divide to a two-dimensional space that distinguished
positions on socio-economic issues and regime matters, respectively. The fact that positions
on redistributive and regime issues were not aligned implies that theories stressing economic
inequality as a driver for regime change are of limited applicability. Instead, the case of
Imperial Germany highlights the importance of cross-cutting non-economic societal
cleavages and the role of societal and political organisations in drawing attention to and
perpetuating these divisions.
Keywords
Dimensionality, political space, roll call votes, Reichstag, Imperial Germany
Political conflict dimensions in Bismarck’s Germany
Germany’s political development before World War I forms the backdrop for a number of
influential theories of democratization (Moore 1966), party development (Michels 1911), as
well as party system creation and change (Lipset and Rokkan 1967). Representing the
paradigmatic case of the ‘late democratizer’ in Western Europe, even contemporary scholars
of democratization feel compelled to demonstrate that their theories can provide an
explanation for the longevity of the Kaiser’s regime (e.g., Acemoglu and Robinson 2006: 67,
200–201; Ansell and Samuels 2014: 28–29, 52–54). This study contributes to our knowledge
about German political development by identifying the dimensionality and content of the
political conflict space during the founding years of the modern German state. It does so by
analysing a novel data set of roll call votes taken in the Reichstag of the North German
Confederation (1867-1871) and during the first seven legislative terms of the Reichstag of the
German Empire (1871-1890). The study period covers Bismarck’s entire term as Chancellor,
which is often seen as a distinct era in the history of Imperial Germany (e.g., Biefang 2012:
17–18). The study complements the early dimensional analysis by Smith and Turner (1981)
of parliamentary voting in Wilhelmine Germany as well as more recent studies by Debus and
Hansen (2010; Hansen and Debus 2012) on parliamentary voting in the Weimar Republic.
However, the study does not only increase our descriptive knowledge about the nature
of political competition in an important historical case, it also contributes to the
methodological debate about the interpretation of roll call analyses and the theoretical
discussion about the causes of regime stability and change. Methodologically, the analysis
illustrates the importance of taking the policy agenda into account when interpreting the
results of a statistical scaling analysis. Apparent changes in the number of dimensions can be
due to the genuine appearance or disappearance of particular conflict constellations as a result
of the entrance or exit of actors or issues, or artificially induced by the legislative agenda
restraining votes to policy issues related to only a subset of the dimensions. In practice, only
substantive knowledge about the content of votes and the historical context of the case can
help distinguish between these two scenarios. This is a particularly important issue when
analysing parliamentary voting in historical cases with comparatively low levels of legislative
productivity and a small number of associated roll call votes. In such circumstances, the
results of scaling analyses are particularly sensitive to agenda effects.
Theoretically, the study results inform the debate on economic inequality as a cause of
democratization (e.g., Acemoglu and Robinson 2006; Ansell and Samuels 2014; Boix 2003).
2
A shared assumption of these political economy approaches to democratisation is that regime
preferences are induced by economic interests. If that was the case, political conflict
dimensions should be aligned with and reducible to divergent socio-economic cleavages.
However, the findings of the scaling analysis show that such an alignment did not generally
exist in 19th
century Germany. For most of the time period studied, positions of party groups
varied along two separate dimensions, one relating to socio-economic issues and one to
regime matters. This separation was mainly due to the emergence of the Centre Party in 1871,
which can ultimately be traced back to a religious cleavage between the Catholic minority
and the Protestant majority in Germany. Given the significant influence of this non-economic
factor in structuring political competition, theories based on economic inequality as primary
causal variable for democratization cannot provide convincing explanations for the longevity
of this authoritarian regime.1
The political system of Imperial Germany
The North German Confederation was established under Prussian leadership in 1867, after
Prussia had defeated Austria in the ‘German War’ of 1866. A few years later, another
victorious war of Prussia, this time against France, paved the way for the accession of the
South German states to the Confederation and for the foundation of the German Empire. The
constitution of the German Empire of 1871 was largely inherited from the North German
Confederation. One of the constitution’s most progressive features was the provision of equal
suffrage for all male citizens of 25 years’ age or older. However, as an intentional
‘corrective’, the constitution banned the payment of allowances to deputies in an effort to
restrict the passive suffrage (Butzer 1999: 29). It also established a strict separation-of-power
system, consciously designed to prevent a parliamentarization of the regime (Rauh 1973: 48).
The Kaiser remained in full control of the military, and the Reichstag’s yearly budget
approval right did not extend to the military budget. The Reichstag shared legislative powers
with the Bundesrat, which represented the governments of the kingdoms and principalities in
this federal monarchy. Contrary to a common misperception, the Reichstag was able to
initiate legislation (Anderson 2000: 10); and in practice, the approval of bills involved
reaching compromises between the Reichstag and the Bundesrat, similar to bicameral
bargaining in current-day legislatures.2
1 For a similar conclusion regarding the value of theories of economic inequality for explaining Argentina’s
democratic breakdown in the 1920s, based on a similar methodological approach, see Alemán and Saiegh’s
(2014). 2 For case study examples, see the descriptions of legislative decision-making in Rauh (1973) or Butzer (1999).
3
By occupying several important posts between 1867 and 1890, Bismarck held a
particularly elevated position in this system. As Chancellor, Bismarck chaired the Bundesrat
and was the only politically responsible member of the imperial government (Reichsleitung).
Government departments were not led by ministers but state secretaries, who were
subordinate to the Chancellor. At the same time, Bismarck was the prime minister and
foreign secretary of Prussia. In this latter function, he commanded Prussia’s votes in the
Bundesrat. Prussia’s votes were sufficient to veto any constitutional change and, together
with the votes of various micro-state enclaves that were completely dependent on Prussia
(Rauh 1973: 60–61), could determine legislative decisions in this body. Although Bismarck
enjoyed strong public support, his power and influence was ultimately dependent on the trust
and backing by the Kaiser and King of Prussia, who could unilaterally appoint and dismiss
the Chancellor or members of the Prussian government.
At the national level, the parties that competed for seats in the Reichstag can be divided
into four broad camps. The liberal camp consisted of the left-leaning German Progress Party
(Deutsche Fortschrittspartei), renamed to German Liberal Party (Deutsch-Freisinnige Partei)
in 1884, and the right-leaning National Liberal Party (Nationalliberale Partei). In contrast to
the Left Liberals, the National Liberals were willing to compromise their political liberal
principles as long as progress was made on establishing a liberal economic order in a unified
nation state. During the first ten years of Bismarck’s reign as Chancellor, the National
Liberals were his primary partner in the Reichstag. The conservative camp consisted of the
Free Conservative Party (Freikonservative Partei) and the German Conservative Party
(Deutsch-Konservative Partei).3 The Free Conservatives, mostly high aristocracy and senior
bureaucrats, were ardent supporters of Bismarck’s policies throughout his term of office
(Stalmann 2000). The German Conservatives, which included more members of the landed
Prussian aristocracy, the infamous Junkers, were more critical of Bismarck’s early policies
regarding national unification and the implementation of largely liberal economic ideas. The
Catholic camp consisted of the Centre Party (Zentrum), often supported by representatives of
regional and national minority groups (i.e. Poles, French, and Guelphs). Some of its leaders
had already been members of the diverse Particularist party group in the Reichstag of the
North German Confederation, but the Centre Party was only formed in 1870. It entered the
Reichstag after the foundation of the German Empire in 1871. Finally, the Socialists
3 The latter adopted the adjective ‘German’ only in 1876, but to avoid confusion with the Free Conservatives
and references to the broader conservative camp encompassing both party groups, the term is also used to refer
to the Conservative Party pre-1876.
4
constitute the last camp. Before their merger in 1875 under the label of Socialist Worker’s
Party (Sozialistische Arbeiterpartei), this camp was composed of two small factions, which
sometimes fiercely opposed each other, the General German Workers’ Association
(Allgemeiner Deutscher Arbeiterverein) and the Social Democratic Worker’s Party
(Sozialdemokratische Arbeiterpartei).
Figure 1 shows the share of Reichstag seats of these party groups over time. Party
groups are roughly ordered along their position on a general liberal-conservative dimension.
Conservative party groups are at the top of the figure, and liberal ones at the bottom. Cells are
shaded according to the share of seats a party group held during a legislative term. Although
the table indicates considerable fractionalization, it also shows that five groups dominated the
Reichstag for most of the period: the two conservative and the two liberal parties, and since
1871, the Centre Party. The table also shows that the Socialists were a fringe group that
played no decisive role in the Reichstag during that period of time.4
At this point, it should be noted that the fragmented multiparty nature of legislative
competition in Germany’s national parliament differed from electoral competition at the
district level. Partly due to Germany’s electoral system with run-off elections in single-
member districts, party political competition at the local level was often limited to two or
three parties; and the identity of these parties often depended on local socio-economic, ethnic,
or religious cleavages. In the most extreme cases, party competition was not only more
limited but almost completely absent, with the election of the candidate of a particular party
being virtually a certainty. The Centre Party was particularly successful in that respect, but a
sizeable number of Polish and German Conservative seats were similarly secure (Schauff
1973: 307). Thus, the analysis presented here describes party political competition at the
national level, resulting from the electoral aggregation of quite disparate political conflict
constellations in electoral districts at the local level.
Figure 1 allows us to make some inferences about the potential influence of different
groups and possible majority constellations at the national level in the Reichstag. First, the
government could never rely on an exclusively conservative majority. The two conservative
party groups reached their highest combined seat share of 34.6% in the Reichstag of the
North German Confederation of 1867. Second, the North German Confederation was
essentially a bipolar system, separating the conservative from the liberal camp, but neither
camp commanded a clear majority on its own. Third, the newly formed Centre Party entered
4 A partial exception is the sixth term of the Reichstag, in which the Socialists held 6 per cent of the mandates.
5
the first Reichstag of the German Empire immediately as the second largest party and from
1874 continuously won about a quarter of the seats. Considering that the minority groups
usually supported the Centre Party, the effective share of Reichstag votes commanded by this
party group was probably closer to one-third. Fourth, the National Liberals were the
dominant party group until 1878. They reached the peak of their parliamentary representation
of 38% during the second Reichstag term, after the Prussian government had thrown its
support behind National Liberal rather than German Conservative candidates during the
election of 1874.
Figure 1 Share of Reichstag seats of party groups by legislative term
Notes: Minorities include French, Polish, and Guelphs. Besides the German Progress Party/German Liberal
Party (Deutsche Fortschrittspartei/Deutsch-Freisinnige Partei), Left Liberals include the Free Association (Freie
Vereinigung, 1867-1871), the German People’s Party (Deutsche Volkspartei, 1871-1890), and the Liberal
Association (Liberale Vereinigung, 1881-1884). Before 1877, Socialists include the General German Workers’
Association (Allgemeiner Deutscher Arbeiterverein) and the Social Democratic Workers’ Party
(Sozialdemokratische Arbeiterpartei). These two groups merged in 1875 to form the Socialist Workers’ Party
(Sozialistische Arbeiterpartei). NDRT = Reichstag of the North German Confederation; RT = Reichstag of the
German Empire. See Table A1 in the online appendix for a more detailed numerical breakdown.
6
However, in 1878, Bismarck dissolved the Reichstag prematurely, blaming the liberal party
groups for the rejection of his first anti-socialist law bill. As a consequence, public support
for liberals of all strands dropped. Furthermore, during the fourth term, the left wing of the
National Liberals seceded from the party group over disagreements regarding the support of
Bismarck’s protectionist and repressive policies. The fifth and sixth Reichstag had the most
left-leaning and oppositional composition. The National Liberals only made significant gains
again in the elections of 1887, after another premature dissolution of the Reichstag by the
government. This time, Bismarck had blamed the Centre Party, the Left Liberals, and
Socialists for rejecting his proposal for another seven-year military budget, claiming that
increases in military spending were necessary to avert an allegedly imminent attack by
France. Forming the so-called ‘Kartell’ electoral coalition with Free and German
Conservatives, National Liberal election candidates enjoyed the support of a united right as
well as the Prussian bureaucracy. However, even this government-friendly alliance broke
apart in 1890 because of its members’ refusal to compromise over the renewed extension of
the anti-socialist law.
Based on the historical record, it is difficult to develop firm expectations about the
dimensionality of the party system’s conflict space or the content of its dimensions,
especially as both might also have changed over time. With respect to its dimensionality,
contemporaries and historians referred to the Centre Party and the National Liberals often as
the ‘middle parties’ (e.g., Nipperdey 2013: 513), implying that they occupied ground in-
between ideological poles. At the same time, they also suggest that Bismarck was able to
govern with varying majorities (e.g., Pflanze 1982: 572), which indicates the existence of a
multi-dimensional conflict space. In general, given the high fragmentation of the party system
and the variety of underlying cleavages, it seems likely that the conflict space was
multidimensional, but expectations about the exact number of dimensions and their content
are hard to discern a priori.
Collection and coding of Reichstag roll call votes
To identify the number and the content of political conflict dimensions in Bismarck’s Reich,
the study conducts a scaling analysis of roll call votes taken in the Reichstag between 1867
and 1890. In the Reichstag, the standard way of voting was to stand up or remain seated. A
proposal for a vote by roll call required the support of at least 50 deputies, which constituted
7
quite a large threshold.5 Before the introduction of voting by division through an amendment
of the rules of procedure in 1874, votes that were too close to call also immediately triggered
a roll call.6 In practice, these thresholds ensured that recorded votes were called on important
and divisive topics that engaged a substantial number of legislators. Possible selection biases
are a fundamental concern for studies of roll call voting (e.g. Carrubba et al. 2006). However,
to the extent that the probability of calling a roll call vote is positively associated with
political conflict, the selective nature of roll call votes actually results in a more informative
sample for the purposes of identifying conflict lines in the legislature; and given the
distribution of mandates across party groups, none of the three main camps (conservatives,
liberals, Catholics) in the Reichstag were prevented from requesting a roll call vote during
any of the legislative terms.
The source for the voting information is the overview of roll call votes in Appendix A
of the ‘General Register for the Stenographical Reports of all Reichstag Sessions from 1867
to 1895’ (Reichstagsbureau 1896). After downloading the digitized version of the General
Register from the website of the Bavarian State Library, optical character recognition (OCR)
software was used to convert the PDF images of the roll call vote appendix into machine-
readable text. To identify the full population of legislators and disambiguate their party group
affiliation and electoral district at the time of a vote, the vote data was linked to biographical
information from the online database Biorab-Kaiserreich, which is hosted by the Centre for
Historical Social Research at the GESIS Leibnitz Institute for the Social Sciences. Before
combining the two datasets through a custom-made record linkage algorithm, information
from both sources was extracted through computer scripts developed in Python. Several steps
in the data collection process required extensive manual review and corrections based on the
consultation of historical primary and secondary sources.7
To assess possible agenda changes in the roll call data over time, the policy area of each
roll call was coded from its brief description in the General Register. The description is
sufficient to identify the general policy area of a proposal, but not for identifying the specific
rationale behind amendments to articles or paragraphs within a proposal. To reliably code
policy content at the level of individual amendments, detailed case knowledge about the
context of hundreds of individual votes would have been required. Absent such knowledge,
5 The total number of Reichstag members was 297 in the North German Confederation (1867-1871), 382 in the
first term of the Reichstag of the German Empire (1871-1874), and 397 in the remaining legislative terms, after
15 new members had joined in 1874 to represent districts of newly annexed Alsace-Lorraine. 6 Stenographische Berichte des Reichstags 1874 II/1, vol. 2 [27], 680ff., 9 April 1874.
7 Further source information, the details of the data extraction and coding process, as well as a description and
discussion of the distribution of votes across types of decisions are given in the online appendix.
8
the coding takes the stated policy goals of the proposal as a whole at face value. However, it
needs to be acknowledged that this coding procedure is likely to understate somewhat the
number of votes that had implications for civil rights and liberties, the relative power of the
Reichstag vis-à-vis other institutions, or the relative power of the Reich vis-à-vis its
component states. In terms of policy categories, the topic coding scheme of the Comparative
Agendas Project was used.8 Despite being widely used in the study of comparative policy-
making, this seems to be the first time that the scheme is applied to a historical case.
In general, the roll call votes in the dataset are mostly about the operation and
organisation of the state (19 per cent), foreign trade regulation (16 per cent), military
organisation and defence spending (13 per cent), law and crime (12 per cent), the regulation
of domestic commerce (11 per cent), civil rights and liberties (10 per cent), macroeconomic
policy (6 per cent), and social welfare (5 per cent). The fact that roll call votes are
concentrated in 8 out of 20 possible policy categories shows how limited the agenda diversity
of the Reichstag was compared to current-day legislatures. The Reichstag was in session only
about 3 to 6 months a year; and the involvement of the state in social and economic matters
only started to build up in the 1880s. Thus, both the capacity and demand for legislative
productivity was comparatively low.
Figure 2 shows that the policy area concentration of roll call votes is even more
pronounced within individual legislative terms. This concentration of the agenda is not a
methodological artefact but reflects a real focus of the Reichstag’s limited attention on a
single or a small number of particularly important and comprehensive pieces of legislation.
These prominent pieces of legislation often involved a large number of amendments decided
through roll call votes. As long as there is a sizable share of roll call votes in policy areas
relating to other conflict dimensions, a strong concentration of attention on a single policy
area is not problematic for identifying the dimensionality of the conflict space or the content
of conflict dimensions. For the purposes of this study, selection effects of the agenda are only
consequential when they focus the attention of the Reichstag on policy areas that relate
exclusively to a subset of the dimensions structuring the conflict space. For example, the
analysis below shows that the scaling results for the second legislative term of the Reichstag
suffer from the effects of a selective agenda; not because its agenda was focused on any
8 For the project, see http://www.comparativeagendas.info; for a list of all topic codes, see
https://dl.dropboxusercontent.com/u/552280/CAPMasterCodebook/MasterCodebookTopics.csv (both last
accessed 11 April 2016).
9
single policy area, but because its agenda was focused on a set of policy areas that all relate
to a single dimension of a multidimensional conflict space.
From a historical perspective, it is interesting that Figure 2 shows a clear pattern in
attention to different policy areas over time. During the so-called ‘liberal era’ of the North
German Confederation (NDRT) and the first three legislative terms of the Reichstag (RT 1 to
RT 3), roll call votes focused to a large extent on measures aimed at clarifying the role and
operation of government institutions, as well as the establishment of a functioning federal
administration. Attention was also devoted to developing a penal code (Law, Crime, and
Family Issues in NDRT and RT 2) and a harmonised court system (RT 2). In contrast, after
what some historians have called the ‘second founding’ of the Reich in 1878 (Barkin 1987;
Gerschenkron 1943), the focus switched from state-building measures to economic and social
policies. In particular, trade tariffs (Foreign Trade in RT 4, 6, and 7), domestic taxes
(Macroeconomic Issues in RT 4, 5, 6, 7), and a new system of social welfare insurance
(Social Welfare in RT 4, 5, and 7) were high on the agenda. Finally, three policy areas were
the subject of roll call votes throughout the period studied here: civil rights and liberties
(especially repressive measures against the Catholic Church during the Culture War and
against Social Democrats after the anti-socialist law was first introduced in 1878), the
organisation and financing of the military, and the regulation of domestic commerce. The fact
that much of the regulation of domestic commerce in the 1880s (RT 5 and 7) was about
‘correcting’ the liberal policies adopted in the North German Confederation in the late 1860s
(NDRT) is another telling sign of the programmatic reorientation of German economic policy
after 1878.
10
Figure 2 Share of roll call votes in different policy areas by legislative term
Notes: Policies were classified according to the classification scheme of the Comparative Agendas Project
(http://www.comparativeagendas.info). NDRT = Reichstag of the North German Confederation; RT = Reichstag
of the German Empire. See Table A2 in the online appendix for a full numerical breakdown.
In total, the data consists of 129,579 potential vote choices. These vote choices relate to 322
roll call votes in seven legislative terms by a total of 2884 members of the Reichstag.
Because the third Reichstag was dissolved prematurely, only nine roll call votes were taken
during that term. This low number of votes precludes a separate scaling analysis. Hence, this
legislative term had to be omitted from the sample. Table 1 shows that, in most other terms,
the number of roll call votes ranged between 30 and 50. Only the Reichstag of the North
German Confederation (NDRT) stands out for its exceptionally large number of roll call
votes (i.e., 81). The large number of roll calls in the NDRT is at least partly a result of the
high legislative productivity of this parliament (Pollmann 1985: 433, 451). It is also
noteworthy that the number of legislators per term is usually substantially higher than the
number of constituent members at the time, indicating substantial turnover of members
within a term.
11
Table 1 Roll call vote characteristics by legislative term